Conceptualizing patient-centered care for substance use disorder treatment: findings from a systematic scoping review

1Substance-related disorders are increasingly considered multifactorial health conditions that require evidence-based and public health responses. Substantial efforts have been made to expand the availability of pharmacological, psychosocial and community-based treatments. In addition, practice-based frameworks, such as trauma-informed and culturally competent, responsive and appropriate care, have also been developed. In spite of these efforts, global estimates suggest that one out of every six people or less in need of substance use disorder treatment receives it. This treatment gap poses a significant public health concern given that treatment engagement (i.e., retention, adherence) is positively associated with improvements in substance use, health and social functioning.

2To understand this gap, a growing body of research has focused on treatment process barriers and facilitators from the perspectives of people using substances. Select recent evidence from across populations and settings reveals similarities in peoples’ experiences. For example, structural barriers include the costs and convenience of treatment, societal stigma and the attitudes and behaviours of health care providers. Research has also shown that people’s preferred treatment goals and outcomes are often incongruent with those of the health care system. Additionally, evidence suggests that people want more opportunities to be involved in the substance use disorder treatment planning process.

3This body of research reveals opportunities to improve the quality of substance use disorder treatment. Examples of existing frameworks include trauma-informed and culturally competent, responsive, and appropriate care. These frameworks emphasize respect for client diversity, an empowerment of people using substances and provider understanding of the varied impact that trauma and ethnicity/culture have on treatment expectations and experiences. In addition, there are emerging interests in the design of patient-centered approaches for substance use disorder treatment (see for example). Patient-centered care (PCC) has been widely recommended to strengthen the quality of health care as it can be universally applied across treatments, settings and providers. This framework challenges traditional approaches to treatment by prioritizing the unique needs of each client and seeking a greater balance in power between the client and provider. In the last two decades, the health and social sciences have expanded the conceptualization of PCC. Although this varies slightly between disciplines and settings, the principles of PCC most frequently include the integration of a holistic or bio-psycho-social approach; an individualized focus on clients’ unique needs, goals and preferences; shared power and responsibility between the client and health care provider as with collaborative care or shared decision-making; and a therapeutic alliance.

4Such varying conceptualizations of PCC have posed challenges to its implementation and measurement of its outcomes. Consequently, an important first step toward designing, implementing and evaluating PCC approaches in substance use disorder treatment is a broad exploration of its orientation and conceptualization in this field. To our knowledge, no such reviews exist, although specific principles of PCC have been empirically studied (see for example). Therefore, the aim of the present scoping review was to systematically explore how the principles of PCC have been defined in substance use disorder treatment. Specifically, this study asked: Which PCC principles have been described in substance use disorder treatment settings?How have these PCC principles been conceptualized?What outcomes of PCC principles have been empirically described?

5Which PCC principles have been described in substance use disorder treatment settings?

6How have these PCC principles been conceptualized?

7What outcomes of PCC principles have been empirically described?

8A scoping review was deemed the most appropriate and feasible synthesis methodology to capture the breadth of existing evidence. This review followed the classic Arksey and O’Malley framework and best practices for conducting and reporting scoping reviews (Additional file 1). The review’s protocol was registered with Open Science Framework (https://osf.io/5swvd/). Full methodological details are available elsewhere and summarized below.

9The search strategy (Additional file 2) was developed as a broad framing of the population (people with substance-related disorders), concept (patient-centered care) and context (health care settings delivering substance use disorder treatment). It was developed in English in Medline (Ovid), refined through extensive consultations with a Health Sciences Librarian and clinical experts (authors SM and SH) and was peer-reviewed. The empirical search for primary studies and previous reviews was conducted in Medline (Ovid), Embase (Ovid), PsycINFO, CINAHL, and ISI Web of Science. The search for grey literature reports and clinical practice guidelines was done in British Columbia Guidelines and Protocols Databases, CPG Infobase, the Registered Nurses’ Association Clinical Practice Guidelines Program, Des Libris, National Guideline Clearinghouse and TRIP.

10Two independent reviewers (author KM and SB) selected references through a two-stage screening process. In the first stage, the reviewers screened the de-duplicated titles/abstracts (85% agreement) according to criteria one through three below. In the second stage, titles/abstracts meeting these initial criteria underwent full-text review (93% agreement) based on the full list of eligibility criteria. Empirical and grey literature references were eligible, if they: Included people with substance-related disorders, including tobacco, alcohol, cannabis, stimulants, opioids or had dual diagnoses.Described patient-centered care (i.e., holistic care, individualized care, shared decision-making, therapeutic alliance), trauma-informed care and/or culturally safe care.Were set in a health care context that delivered substance use disorder treatment. This included inpatient (e.g., hospital, residential treatment) or outpatient (e.g., emergency department, primary care, community-based program) settings. This excluded criminal justice settings and self-help models.Were published between 1 January 1960 and 1 July 2018 in English, French, Spanish, Italian or Portuguese.Provided an operational or conceptual definition of the patient-centered care approach.Empirical quantitative references observed at least one patient outcome (e.g., substance use, health) or treatment process outcome (e.g., treatment engagement, treatment satisfaction).

11Included people with substance-related disorders, including tobacco, alcohol, cannabis, stimulants, opioids or had dual diagnoses.

12Described patient-centered care (i.e., holistic care, individualized care, shared decision-making, therapeutic alliance), trauma-informed care and/or culturally safe care.

13Were set in a health care context that delivered substance use disorder treatment. This included inpatient (e.g., hospital, residential treatment) or outpatient (e.g., emergency department, primary care, community-based program) settings. This excluded criminal justice settings and self-help models.

14Were published between 1 January 1960 and 1 July 2018 in English, French, Spanish, Italian or Portuguese.

15Provided an operational or conceptual definition of the patient-centered care approach.

16Empirical quantitative references observed at least one patient outcome (e.g., substance use, health) or treatment process outcome (e.g., treatment engagement, treatment satisfaction).

17Study screening (including de-duplication) and data extraction was done in DistillerSR. The data charting form (Additional file 3) was used to capture the characteristics of each reference and to identify which principle(s) of PCC were described (objective 1). This form was piloted with the first five empirical and grey literature references. Author KM led data extraction and author SB checked extraction of the PCC principles (94% agreement).

18A descriptive overview (including tabular and graphical summaries) of extracted data was completed. In addition, a directed content analysis was performed on the defining characteristics of PCC principles and their outcomes (objectives 2 and 3). This systematic method is particularly beneficial when there exists theories or frameworks (i.e., for patient-centered care) that can guide coding and analysis and can be used to determine patterns and relationships between the content coded. For this analysis, a coding guide (Additional file 4) was developed to identify categories (e.g., therapeutic alliance), sub-categories (e.g., defining characteristics, outcomes associated with) and codes for the defining characteristics (e.g., non-judgment) and outcomes (e.g., frequency of substance use) of each PCC-principle. This guide was developed in iterative stages through consultation with the team’s knowledge users and the initial data extraction process; these codes and categories were then broadly operationalized according to existing PCC frameworks. Directed content analysis also allows new evidence to emerge via open codes, which are used to label content that is unique to predetermined codes. Open coding was used within each of the broader categories (e.g., therapeutic alliance/defining characteristics) and also to identify antecedents to PCC.

19Author KM led initial coding; after which, the categories, subcategories, codes and content were reviewed with the team for trustworthiness and for further analysis (e.g., integration, collapsing, or expanding categories). Since the content coded across categories was not mutually exclusive (i.e., a reference could have more than one principle of PCC or more than one category of outcomes coded), data reported are the number of references (i.e., sources) coded at each category, instead of the number of times each reference is coded. The directed content analysis was carried out in NVivo (version 11 for Mac).

20After de-duplication, 2951 unique references underwent title/abstract screening. Of these, 395 were assessed during the full-text review, with 149 references included (Fig. 1; Additional files 5 and 6 for Excluded References and Detailed Extracted Data). Table 1 provides the characteristics of the eligible references. Approximately 50% of the references were empirical quantitative papers published in the last decade, over two-thirds were based in North America and all but three of the references were published in English. The targeted population was primarily adults receiving substance use disorder treatment. Nearly two-thirds of the references were in outpatient settings and delivered psychosocial treatments. Regarding the principles of PCC, 63 (42.3%) references described more than one PCC principle, and therapeutic alliance was the most frequently described (n = 107; 71.8%).

21Tables 2, 3, 4, 5, 6 and 7 summarize the results of the directed content analysis of defining characteristics for each PCC principle, and include representative examples of the content coded. For holistic care (Table 2), the sub-categories converged in their aim to provide “wraparound services that meet clients' needs at a given point in time”. For most of these references, this included an integrated delivery (n = 25) or the coordination (n = 15) of additional health (e.g., primary care, specialist care, nutrition, exercise) and psychosocial supports (e.g., housing, financial, legal, family) within a substance use disorder treatment setting. Gender-responsive services (n = 9) described an integrated approach, wherein women’s health, substance use and psychosocial treatment needs were comprehensively addressed. Finally, four references described the delivery of substance use disorder treatment within a primary care or hospital setting.

22Individualized care (Table 3) was defined by health care providers’ efforts to understand clients’ unique needs, preferences, and expectations. The first of such efforts was the use of individualized assessments and treatment plans, both at entry and throughout treatment (n = 29). Eight of these references (8/26 = 30.8%) used a specific tool (e.g., Goals of Treatment Questionnaire) in these assessments. More frequently, a general process of assessment was described, whereby health care providers took time to understand the “main problems to be addressed, what actions and resources were needed, who is responsible and timeframes for action and review”. The second defining sub-category described efforts to deliver treatment by “fit [ting] services to the individual, based on an ongoing assessment of that person’s needs and level of functioning” (n = 24). This often included presenting clients with a range of treatment options that responded to those assessed needs and preferences. Examples of treatment options included group or individual counseling, medication options (when there was more than one), the schedule and frequency of visits, and location of visits. This was sometimes referred to as “treatment-matching” (n = 9) or “as-needed dosing” (n = 3).

23Shared decision-making (Table 4) was defined in the first sub-category by activities or strategies whereby the client and provider engaged in dialogue to reach a mutual decision on the best course of treatment (n = 31), including choice of the intervention, its frequency, duration, and follow-up plans. Here, health care providers elicited clients’ preferences and needs, presented information on the available treatment options, and then the client and provider “negotiated dialogues towards a mutually agreed upon destination”. In the second sub-category, decision-making was referred to as an autonomous and client-led approach (n = 17). Clients were described as having responsibility and control over their treatment decisions, including the frequency of counseling (e.g.), choice of medications or behavioural interventions, and transition plans (e.g.). These two defining categories were similar in their empowering view of clients as an “integral partner, rather than passive or compliant recipient, of a treatment program”. They diverged in the extent of autonomy that the client had and in their emphasis on the dialogue process.

24Therapeutic alliance was defined by relationships that were non-judgmental, respectful and accepting (n = 37) and/or as empathic, understanding, warm and kind (n = 32). While these defining categories reflected distinct relational qualities, there was substantial overlap between them (n = 26 references coded at both), as shown in the representative examples in Table 3. In addition to these characteristics, 56 (52.3%) references defined therapeutic alliance according to empirically-based measures, such as the widely used Working Alliance Inventory. Among these references, therapeutic alliance was commonly explored as a predictor or mediating variable of substance use and treatment engagement outcomes.

25Trauma-informed care was defined according to existing theoretical and clinical practice frameworks (n = 6), such as Seeking Safety (n = 2) and Harris & Fallot’s trauma theory (n = 2). Open coding captured additional defining features, including understanding the effects of trauma (n = 3) and avoiding re-traumatization (n = 1). Defining characteristics of culturally-safe care included open codes for adapting care plans according to culturally-relevant preferences of clients (n = 7), inquiring about the health and healing beliefs of clients (n = 2), and health care providers’ reflection of their personal beliefs and biases (n = 2).

26A total of 103 (69%) references were identified as describing or exploring at least one of the predetermined categories of outcomes. The sankey diagram (Fig. 2) displays nodes for each principle of PCC, as well as the categories and sub-categories of coded outcomes. The width of each flow represents, among those references that identified outcomes, the relative distribution of which PCC principles were studied or described, and in association with which outcome categories and sub-categories. As shown in Fig. 2, therapeutic alliance and shared decision-making contributed a higher number of references with identified outcomes. Within each of the PCC principles, substance use (n = 52/103; 50.5%) and treatment engagement outcomes (n = 50/103; 48.5%) were the most frequently coded categories, followed by health and psychosocial outcomes (n = 40/103; 38.8%), and patient-reported experiences (n = 17/103; 16.5%). Within each of the broader outcome categories, the most frequent subcategories included the number of days of substance use, number of visits or sessions attended, physical and mental health symptoms, perceived self-efficacy, and treatment satisfaction.

27During the directed content analysis, open codes were used to distinguish antecedents contributing to, or strengthening the implementation of PCC from those that were described as outcomes. Since this was not part of the original review design, these factors were analyzed inductively and based on their within-category content. A total of 75 (50.3%) references were coded for describing such contributors; the emergent categories and sub-categories are displayed in Table 8. Organizational values, policies, and procedures (n = 42; 56.0%) clustered around six features. These included the skills and training of providers (e.g., case management, multicultural competence) and environments that were safe, stable, and social. Clinical approaches (n = 49; 65.3%) also emerged as contributing to therapeutic alliance (n = 43; e.g., communication style, building trust), shared decision-making (n = 30; e.g., appropriate information sharing; empowering approach), or individualized care (n = 6; e.g., encouraging client input).

28To our knowledge, this scoping review is the first to undertake a systematic synthesis of PCC in substance use disorder treatment settings. To strengthen the breadth and specificity of this review, existing frameworks of PCC from other disciplines were used to guide the search strategy and data charting methods, and the directed content analysis allowed population and context specific nuances to be identified. The findings suggested that few references had examined all four principles of PCC, although 42% described more than one PCC principle. The most frequent principle identified was therapeutic alliance and the most frequent outcomes measured included substance use and treatment engagement. The findings contribute evidence that can be used to support a comprehensive and evidence-based conceptualization of PCC with implications for its implementation and evaluation.

29The first objective was to determine which PCC principles have been described in substance use disorder treatment settings, and the results revealed that therapeutic alliance was the most frequently described principle. The first plausible explanation for this is the longstanding tradition of therapeutic alliance in psychotherapeutic research and practice. In the present review, two-thirds of the references offered primarily psychosocial treatments (e.g., cognitive behavioral therapy) for substance-related disorders. In this discipline, therapeutic alliance receives significant attention given its importance in predicting counseling outcomes. In the references that described therapeutic alliance, over 50% were empirical quantitative papers and conceptualized therapeutic alliance according to client, provider, or observer-rated empirical measures, such as the Working Alliance Inventory (WAI). Thus, this longstanding tradition to examine the extent of therapeutic alliance likely contributed to the high number of references in the present review that described this PCC-principle.

30Our search also yielded references that delivered additional treatments (e.g., pharmacological treatments alone or combined with psychosocial) in alternative settings (e.g., residential detoxification programs, harm reduction services), which provided an opportunity to determine that non-judgment, respect, empathy, and understanding were also common characteristics of therapeutic alliance. While respect has been described in broader conceptual analyses of PCC, information regarding why these attributes were important among people with substance use strengthens its interpretation in this context. Examples of these reasons included clients’ safety; to gain clients’ trust; and to reduce stigma. These defining attributes are especially salient when considering that experiences of stigma are common among people with substance-related disorders and have been identified as barriers to treatment.

31These relational characteristics also intersected with shared decision-making, such that the analysis of antecedents to PCC revealed that respectful and understanding relationships promoted shared decision-making. The reciprocal was also found, whereby collaborative approaches strengthened therapeutic alliance. These antecedents give more depth to our finding that the defining characteristics of shared decision-making denoted an underlying philosophy of respect towards clients as “integral … rather than passive” partners in the treatment process. The first defining characteristic emphasized a joint decision-making process. This category primarily described a process of dialogue and discussion that granted clients a more active role in the decision-making process and facilitated the health care provider’s understanding of clients’ needs and expectations. This view of shared decision-making resembles those of the broader PCC-frameworks that have conceptualized this principle as “sharing power and responsibility”, “finding common ground” and also more recent proposals for the clinical practice of shared decision-making. However, the second defining characteristic emphasized a fully autonomous decision-making process, which is more closely aligned with other frameworks’ notion of “empowering care”. Those existing frameworks describe autonomy, self-confidence and self-determination as core characteristics of this principle. While, there is evidence that increasingly recognizes clients’ preferences to be more actively involved in substance use disorder treatment decision-making, further work might explore the circumstances under which a deliberative or autonomous decision-making process is more suitable from the client and health care provider’s perspective.

32The integration of shared decision-making practices often presumed an individualized care approach, such that the process of dialogue involved discussion of clients’ unique needs, circumstances, traditions and preferences. One study to highlight is that of Joosten et al. who developed and tested the effectiveness of a shared decision-making intervention in an inpatient treatment setting. Their intervention relied on an individualized assessment of clients’ needs and goals (via the Camberwell Assessment of Need). The client and clinician then discussed their independent ranking of the priority of these goals and adapted treatment accordingly. In this review, individualized care did not always include shared decision-making however; it was also described by several treatment matching approaches, such as as-needed-dosing. Regardless of the specific design chosen, these findings imply that comprehensive assessments and flexibility in service design and delivery (at a clinical and organizational level) supported individualized care.

33Thus, individualized needs assessment and treatment delivery overlapped with holistic, trauma-informed and culturally competent, responsive, and appropriate care with respect to their common goal to provide comprehensive and flexible care, adapted to client-identified needs and values. Under ideal circumstances, such consideration would be facilitated by an assessment of clients’ bio-psycho-social needs, which are often inextricable from their cultural context and the pervasive impacts of structural and interpersonal trauma. In the present review, the defining characteristics of these principles included specific practices adopted by health care providers (e.g., comprehensive needs assessments, avoiding re-traumatization). However, the inductive analysis of antecedents to PCC revealed that both the system (e.g., a vision of shared governance; safety and stability of treatment setting; flexibility of service provision) and the health care provider (e.g., communication style) play a conjoint role in the successful implementation of PCC principles. For instance, a physician’s endeavor to adopt shared decision-making practices in the prescription of opioid substitution treatment will require a health care system that has implemented evidence-based treatment options and policies that support client-provider collaboration (e.g., flexibility around dosing schedules, frequency of visits, etc). Thus, a good starting point for moving PCC into the realm of evidence-based practice in substance use disorder treatment is a consideration of potential barriers to its implementation from the client, provider and system’s perspectives.

34While factors supporting PCC have been relatively consistent across broader concept analyses, there has been less agreement on what consequences or outcomes of PCC can be expected and thus, measured. Examples of such outcomes have included consultation processes (e.g., communication skills, quality of care, treatment satisfaction), health behaviours (e.g., service utilization, adherence to treatment plans), health outcomes and patient-reported outcomes. It has also been proposed that some of these outcomes are likely more intermediate (i.e., perceived quality of care, satisfaction, consultation process outcomes), while others more distal (i.e., health behaviours and health outcomes).

35In the present review, substance use and treatment engagement outcomes were the most frequently investigated, regardless of PCC principle. This might have been influenced by the high frequency of references exploring therapeutic alliance, half of which related the WAI with the number of days of substance use or number of counseling sessions. However, it might also reflect a common assumption that the goal of any substance use disorder treatment is to reduce the severity of use. A continued emphasis on substance use outcomes neglects that the stated goal of PCC is to improve the treatment process. It is also not congruent with prior research demonstrating that clients’ goals extend to other domains (e.g., health, housing, family relationships) and emerging recommendations to integrate patient-centered or patient-reported measures in substance use disorder treatment. Thus, future PCC research in substance use disorder treatment will be strengthened through choices of measures that reflect these goals.

36This review was a necessary first step towards conceptualizing PCC for substance use disorder treatment. Although scoping reviews typically take a broad framing of the population, concept and context, this resulted in a high number of false positives and posed several challenges to the synthesis and to teasing apart potential differences in treatment type and setting. While efforts were made to overcome these challenges (i.e., substantial resources were devoted to reaching adequate inter-rater agreement and carrying out the directed content analysis), there are further limitations to bear in mind. First, we were unable to carry out a comprehensive grey literature search in international search databases, other than TRIP, and thus, might not have adequately captured grey literature reports of the implementation of PCC in settings outside of North America. In addition, our search strategy was developed in English and eligibility was limited to references published in English, French, Spanish, Italian or Portuguese. This might have influenced the comprehensiveness and international breadth of the search and thus, inflating the number of references from North America.

37The present scoping review synthesized existing empirical and grey literature as a necessary first step to explore which PCC principles have been described and how they have defined and measured for people engaging in substance use disorder treatment. The directed content analysis revealed population and context-specific evidence regarding the defining characteristics. These results can be used to support the implementation and evaluation of PCC. The results also identify future directions for research, including potential measures of PCC and its associated outcomes.